DIY Audio Projects Forum

KT120 Oddblocks
Page 12 of 49

Author:  DrRoyal [ 20 Jan 2014, 02:37 ]
Post subject:  Re: KT120 Oddblocks


Do you have somebody experience with Hammond 1650RP for KT120
or Plitron 70W-80W

I spent a lot of time but didn't find one opinion about toroid transformers just only on some forums info missing bottom in sound. But some forums sing a song about Plitron.

I understand , Edcor - this is something between. But between what in KT120 monoblock :) ?
Can I get better result with different iron and where this way ?

Thank you for help


Author:  poty [ 20 Jan 2014, 06:31 ]
Post subject:  Re: KT120 Oddblocks

Well, I'd wait for Bruce's answer. Meanwhile...
You won't find one truth for all situations for toroid transformers, because there is no such truth. I know designs that work very well with toroids, also know many designs in which I'd prefer some classic ways. SIPP is very promising in the field of using the toroids, because it has constant current source and balancing circuit, which makes both halves' currents equal (the main problem with the toroids - they do not have any gap, so any constant magnetic bias shortens the power they can transfer).
Specifically about the chosen transformers:
Hammond is a classic one. ... has almost the tested 3,5k primary impedance ... rated for the currents involved... has choice of the load impedances. Comparing to Edcor - the 1650 is less capable in the low frequencies, but I'd not trust the Edcor digits, because they do not disclose any specifications of their testing (for what power they tested the bandwidth), on the other hand the Hammond in the design would be hardly used in the half of power - increasing the rated bandwidth significantly.
Plitron is a boutique one. There is the name of the designer which could bring some trust. :) On the other hand the primary impedance is not too high (most close to Bruce's is PAT-4004-01 2756Ohm 70W), but no current rating and only one load impedance. Much more info about bandwidth (than for Edcor), but with my limited English I cannot understand all of them. If the out of table line: "-3dB Power bandwidth starts at 22.7Hz" means that at half of the 70W the low-freq margin is 22.7Hz - then it is very good value!
To talk about common case of using different iron - I think there is a way to improvements (for example I use a local guy manufactured transformer with very good results), but this is a tough way. There should be some "sweet point" in primary impedance led to your like of power vs distortions. There should be bandwidth considerations. Weight. Size. Rather expensive testing...

Author:  gofar99 [ 20 Jan 2014, 09:34 ]
Post subject:  Re: KT120 Oddblocks

Hi, Part of DIY is trying things that appeal to the builder. So feel free to try other output trannies. However.... there is no guarantee that they will be as good. I have measured the Edcors in the amps and they will deliver full power at any frequency between 20HZ and 20K. That is what they spec for. My experience is that since you never run full power at any frequency a measurement at about 10 watts is more meaningful. At that level they will (in these amps) deliver response within 0.2db from 15 HZ (some will go as low as 10-12 HZ) to past 40K HZ. They actually have a slight rise (about 1-1.5 db) in the 40K to 80K HZ range before falling rapidly after that. They are several db down by 90K. The rise is in the self resonant frequency range and is the reason for the very modest amount of NFB. It insures stability and prevents ringing. Usually the amps can run without it, but I can't predict everything that might be used as speakers and some present difficult loads. BTW the amps are quite happy with ESLs. I use them all the time as my mains.

So while other trannies may work and perhaps even be better, I don't recommend them. With the bandwidth and low distortion levels using the Edcors, I doubt the extra cost and risks would be worth it.

BTW, I get nothing back from Edcor for using their products. I pay for them (as does Oddwatt Audio) at the rate based on quantity ordered like everyone else.

Good listening

Author:  gofar99 [ 20 Jan 2014, 09:42 ]
Post subject:  Re: KT120 Oddblocks

Hi Everyone, A project update. In the KT120 amplifiers based on some beta tester information I have determined that while the sound of the NOS 12SL7s are really lovely and accurate if you want a bit more brilliance or what ever you want to call what seems like more detail you can do a simple rewire and use the Sovtek 6SL7s. Just series a 20 ohm 10 watt resistor in series with the 6SL7 tube heater (not the KT120s) and you are good to go. I put the resistors on the more negative side of the DC string but it probably would not make a significant difference on the more positive side either. This behavior is similar to the KT88 amps with the differences between NOS Philips 5751s and Sovtek 5751s. A different flavor if you choose between them.

Good listening

Author:  DrRoyal [ 20 Jan 2014, 12:15 ]
Post subject:  Re: KT120 Oddblocks

Thank you for respond. I'm just looking for some improvement in sound. I've found info about Plitron (toroid) in 300B SE and this guy very happy with toroid.
It still very expensive experiment :) but just interesting to know Bruce opinion about it.

Also somebody can tried already, right ?

I'm just looking for next project and if I decided to stay with Edcor - time to order set for KT120.


Author:  DrRoyal [ 20 Jan 2014, 23:11 ]
Post subject:  Re: KT120 Oddblocks


Thank you Bruce for update. But in this case no reason to have heater 12V. Will be enough 6V for 2xKT120 and 6SL7. Another question, for power transformer I need something about 5.45V-5.5V and 5A and 360V 0.5A.

Bruce, from your experience (I'm sure you are familiar with Edcor's variety) what will be suitable in this case?


Author:  gofar99 [ 24 Jan 2014, 09:21 ]
Post subject:  Re: KT120 Oddblocks

Hi, I don't believe they have any quite like that. To get 6VDC for the heaters you will need a transformer that starts with something that would account for the losses in the rectification process. It is possible to calculate it, but my experience has been that while on occasion the results are close in practice they are often not. What I would do in this case is use a standard 6 VAC winding, rectify and filter it (at least 20Kuf) and run it to the tubes and see how close it is. I would estimate with a good bridge rectifier and low impedance transformer windings it would be about 0.5 to 1.5 volts high under load. You can use the method employed in the early amps of putting series rectifiers in series with the output to bring it down. Each rectifier will drop the voltage by around 0.5 to 0.7 volts. Be sure to use high current ones and mount them where they can breathe as they will get rather hot. At 5 amps each will have to dissipate about 3.5 watts. I would use screw terminals to hold them in place as the leads can actually get hot enough to desolder themselves if you use standard solder. I had that occur in a very early prototype (way back in 2008). Embarrassing to have the sam suddenly go quiet and have parts fall out the bottom. :blush:

Good listening

Author:  gofar99 [ 21 Apr 2014, 17:41 ]
Post subject:  Re: KT120 Oddblocks

Hi Everyone, We had a beta tester compare a few of the KT120 amps a little while ago. The ones using NOS 12SL7s (under Western Electric contract from 1946) were noticeably warmer than ones using new production Sovtek 6SL7s. Each has merit. (BTW the difference is much smaller with the KT88 amps and neither XSL7 is really that much different from the other in sound) You can make the swap easily if you use 12 volt heaters. Just put a 20 ohm 10 watt resistor in series with the heater going to the 6SL7. If you want to make it compatible for both types a simple switch that shorts out the resistor in one position will work just fine. Just remember which position is shorted and which is not unless you like exciting flare ups. :hot:

Good listening

Author:  reubster [ 22 Jul 2014, 21:54 ]
Post subject:  Re: KT120 Oddblocks

Hi Y'all
First post, I love the site.
Over the years I've spent a lot of time building and modding solid state hifi gear and plenty of time building and modding valve based guitar amplifiers.
It's time to bring it all together and build a new hifi amp, and after much research it will be the Odd Block KT120's

A couple of questions:

Within this discussion, the power transformers seem to produce a B+ around 450 - 460V
OT = 3.5K, Current is ~ 150ma per tube, Voltage across CCS is ~40V
By my "hack" calcs, this produces a best case dissipation of (450 - 40) x 0.150 = 61.5 watts
It seems very high (ie at the KT120's limit)....Am I doing the math wrong (I normally work with AB PP & not CCS) ?

I'm also curious to understand how the OT Primary selection ended up at 3.5K when the KT88 version is 8K.
It seems a big change in impedance and this worries me with the high dissipation?

Finally, in class AB PP circuits the power transformer seems to have an influence on "quality" of audio.
eg higher VA rating = Lower power supply impedance, so I usually overspec my PTs for this reason.
For the Odd Blocks operating in class A PP with a CCS, it seems the current will always remain constant.
Therefore has anyone found any advantage to overspecing the PT (apart from heat)?

Many thanks to Bruce for such an interesting design.


Author:  gofar99 [ 22 Jul 2014, 22:59 ]
Post subject:  Re: KT120 Oddblocks

Hi, I believe you will find the amps quite good. In the latest iteration I reduced the idle current to 290ma per pair from was was 310-315. Just to be safe. I have been running a pair at the higher value for over two years now and there have been no problems. What needs to be remembered is the idle current measured at the cathode is the sum of the anode and screen currents. The combined dissipation rating is 68 watts. I suspect you already know, but others might not...that the dissipation in a class A amp is highest at idle so the actual dissipation will be a little lower depending on how much power you are delivering to the speakers. The load impedance was determined to be the point at which the power transfer was maximized with low distortion. I considered both higher and lover values. Lower would transfer a little more power but have more distortion. Higher would give lower distortion at the expense of power. In the KT88 amps I went for a higher impedance as I wanted the distortion to be quite low (it is) and was willing to give up a few watts. You could run them at 6500 ohms and get about 2-3 more watts but the distortion would nearly double. These amplifiers run with nearly zero negative feed back so the added distortion would not be fixed there. That gets us into a controversial area. I don't care for the sound of amplifiers with a lot of NFB. To me it seems to alter the sound. The amount I use (between 2 and 3 db depending on the amp) is just enough to insure stability under difficult loads and in most cases can be disabled with no ill effects.

As you noted there is no particular reason to increase the transformer rating. It is well over what is needed for operation. The filters are rather generous and will stiffen the B+ any how and provide a sufficiently low impedance return path for any output up to max.

BTW the two resistors in the LM317 circuit are 6R8 and 14R for the lower idle current. They were 6R8 and 10R originally. The loss in output from the change is slight and you will never notice it.

Good listening

Page 12 of 49 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group